L2 Planet Issue #27
In this Issue of L2 Planet, we focused on the developments regarding Fuel Network, Taiko, Mantle, Base, Arbitrum, and zkSync Ecosystem.
FuelProver
Fuel Network has introduced Hybrid Proving, an essential technique bridging the gap between Optimistic and zkRollups.
With Hybrid Proving called FuelProver, we expect a decrease in withdrawal time from typical 7 days to approximately 8 hours. I can hear you saying that how can it be possible?
There will be two provers, the chain will be operating as standard Optimistic Rollup when there isn’t any problem. If any problem occurs and an invalid block is generated, ZK Dispute Prover will move in and prove the block's validity. ZK Prover architecture is agnostic and additive which means it can support all the latest general-purpose provers such as zkSync, Risc Zero, etc.
This idea parallels with Multi-Prover approach of Vitalik Buterin and the work of Metis Foundation’s Hybrid Rollup design. Seems like we will see that Rollups will evolve more in the future!
Taiko Prover Design
Ethereum-equivalent zkRollup solution, Taiko, is waiting for your feedback for its proving design! The team released two testnets with two different proving designs, Grímsvötn L2 testnet (Alpha-3) and Eldfell L3 testnet (Alpha-4).
In the first design, Grímsvötn L2 testnet, EIP-1559 style block proving method was used. The most efficient prover wins logic applied. Everyone can be prover and leave whenever they want.
In the trade-off between the lowest cost and fastest withdrawal for users in a given time period, the most efficient is chosen. The reward is calculated based on the historical and adjusted every time period. This is the part where the EIP-1559 mechanism is applied.
In the second design, Eldfell L3 testnet, staking based proving design was considered. In short terms, for each block, a random prover is selected based on the stake amount of the prover and the expected reward. Anyone can become a prover and leave whenever they want. Stakes can be withdrawn after 1 week.
For each block, the prover is randomly selected from among the 32 provers with the largest weight. Weight depends on a number of parameters. These parameters are the locked Taiko token amount A, the expected reward amount per gas R, and the number of blocks that the prover can prove in parallel is C. If the assigned prover fails to prove the block within a certain time frame, it gets a slash. In this case, any prover can provide proof and prove the block. These blocks are called "open blocks".
In this design, Taiko is winking at a possible token and airdrop :)
Mantle Mainnet
Mantle, an Optimistic Chain, after the six-month developments, has launched its mainnet!
Total asset nearly $60m bridged to Mantle mainnet. More than $20m of this value is ETH. USDT, USDC, and LUSD follow it respectively.
In the testnet phase, Mantle has processed 14 million+ tx, saw 140k+ contracts deployed, welcomed 48k+ developers, connected to 690k+ unique wallet addresses, and maintained a monthly active wallet addresses count of 157k+. The testnet phase helped Mantle to remove potential risks and vulnerabilities.
Current state, fraud proofs are in development and not active on Mantle Mainnet. This means the system permits invalid state roots. Also, data is stored in the MantleDA which is forked from EigenDA with significant modifications such as removal of slashing conditions. Users have to trust that the data is held externally by the team in order to establish the chain’s state.
The system contracts also have upgradeability. Contract owners can change it immediately without notice. In the case of sequencer failure, users be able to submit transactions via L1 but can’t force them. The sequencer can censor the entire queue. In the case of proposer failure, withdrawals are frozen because only whitelisted proposers can publish L2 state roots on L1.
We believe Mantle will solve its security problems in a short time.
Base Mainnet
🟡 Onchain Summer has started! Coinbase’s Base Network, the Optimistic Rollup that leverages Optimism's OP Stack, is now live on public Mainnet.
While only a few days passed after the mainnet launch, more than $158m was bridged to BASE. More than $130m of these values is ETH. Then USDC, DAI, cbETH, and BAL follow it respectively.
%54 of total users deposited between 0-0.01 ETH to bridge contract. Seems like airdrop hunters be aware of the $BASE token airdrop potential.
Arbitrum BOLD
After months of research and development, Arbitrum has announced BOLD, the dispute protocol that will enable permissionless validation.
BOLD will significantly improve decentralization in the Arbitrum Nitro, Nova, and Orbiter chains. As it is known, fraud-proof production in Arbitrum is subject to permission and can only be produced by whitelisted actors. Because a bad actor can prolong L1 approval and withdrawal times by constantly spending money. This is undesirable. These are called Delay Attacks and were previously mentioned by Ed Felten of the Arbitrum team.
For BOLD, Arbitrum Nitro, Nova, and Orbiter, it guarantees L1 confirmation time to an upper limit and ensures chain liveness. The team also shared a sneak peek from the Challenge Visualizer and API developed by Preston Van Loon.
The next step will be to share instructions for running an Arbitrum Nitro devnet with BOLD challenges enabled in the coming weeks.
Is it Plonky2 or Boojum?
As you remember, the Zksync team recently introduced the new proofing system, Boojum.
According to Alex, Co-founder of zkSync, Boojum was inspired by Polygon's Snarky2. But, the Polygon Zero team suggested that it was copy-pasted without any attribution rather than inspiration. The Polygon team also says that the statement that Boojum is 10x faster than Plonky2 is false. In Celer benchmarks, Plonky2 has 16x more constraints than Boojum, so Boojum isn't actually faster. It is even stated that Starky, which was developed by the Polygon team, is much faster than Boojum. The team said that this behavior does not comply with ethical values and announced that Plonky3 will be announced soon
.
After the Polygon’s statements, Alex said that these are untrue accusations and explained why. Both Plonk2 and Boojum are implementations of Redshift which was developed by Matter Labs 3 years before the Plonky2 paper is one. Second, only %5 of Boojum code is based on the code of Plonky2. Also, Polygon didn’t give credit when they used Redshift but Boojum provided clear attribution in line #1 of the main file of the module for reused code.
So, what do you think of these expressions?
Spectating Corner
Reading Corner
Eldfell L3 update: Prover slashing explained, Taiko
An incomplete guide to Folding: Nova, Sangria, SuperNova, HyperNova, Protostar, Lisa A.
BOLD, Permissionless Validation for Arbitrum Chains, Offchain Labs
Tendermint for Starknet, ilia
How I learn Zero Knowledge, Lauri Peltonen
That’s all from L2 Planet for now, hope to see you in 15 days :)